On Monday, Thrane & Thrane announced an agreement with KVH Industries to distribute KVH’s TracPhone V7 mini-VSAT product. At first sight, this seems to be an expression of confidence in the potential of maritime VSATs, with Thrane noting that “the market for Ku-band equipment for maritime satellite communication is expanding strongly” by about 1500 terminals per year. However, Thrane’s original plan was to pursue development of its own maritime Ku-band VSAT product and service, the SAILOR 900, announced last June, which has now been cancelled.
The Ku-band maritime VSAT market has grown dramatically in recent years, taking market share amongst high end users, who can afford the expensive terminals (typically $50,000+) and need the flat-rate, always-on connectivity that Inmarsat cannot provide with its L-band satellites. However, the Achilles heel of Ku-band systems has always been coverage: as Connexion-by-Boeing found out in 2006, the cost of leasing a global Ku-band network far outweighs the near term revenue that can be gained in what is a very fragmented business, with the largest players (Vizada/Marlink, SeaMobile/MTN, Caprock and ShipEquip) each having a market share of less than 20%. As a result, providers have to date offered C-band systems to customers requiring global coverage, and focused Ku-band solutions on intra-regional users.
KVH and Viasat have promised to change this model with their mini-VSAT network, building out global Ku-band coverage for both aeronautical and maritime applications, along similar lines to the concept envisaged by Connexion (though with smaller, cheaper terminals costing around $30,000). Thrane’s decision to join with KVH is an acknowledgement that even a company such as Thrane, with it’s very strong maritime brand, is unable to justify the costs of establishing global Ku-band coverage on its own. The question now is whether KVH and Viasat can afford the investment needed to complete their 2009 coverage expansion plans.
Orders have remained strong for maritime VSATs, even in recent weeks, but in the current financial climate, maritime VSAT looks far more vulnerable than Inmarsat to ship operators seeking to cut back their communications spending: it generally has a high fixed cost every month (compared to Inmarsat’s pay-by-the-use) and is usually sold on the basis of getting “ten times more data for double Inmarsat’s cost”, not on the basis of saving money over current expenditures. Most importantly, in many cases it is seen primarily as an investment in crew welfare: when there are few jobs going either at sea or on land, crew retention in 2009 will be much less of a problem than in 2008. As a result, Thrane looks amply justified in getting cold feet over pursuing its own VSAT product.
Last week, the FCC awarded ICO its ATC license, with no requirement to enter into a ground spare contract at this stage. Indeed ICO will only have to demonstrate 30 days before commencing ATC service that “firm arrangements are in place to meet the spare satellite requirement” within 1 year of service initiation. Of course, since a satellite would normally take around 2 years to construct, ICO would have to enter into a ground spare contract well before that 30 day notice period. However, ICO has gained considerably more flexibility to keep its expenditures to a minimum, while pursuing resolution of its litigation with Boeing and waiting for a more favorable economic climate in which to launch commercial service.
With this ruling, along with Globalstar’s ATC license grant last year, the FCC has shown a desire to be flexible in its interpretation of the ATC rules, so as to ensure that ATC deployments do eventually take place. Arguably, this flexibility has advantaged operators such as ICO and Globalstar, over TerreStar and Skyterra (formerly MSV), who have committed to considerably greater capital expenditure, based at least to some degree on a more cautious interpretation of the ATC requirements. Most notably, TerreStar is already well advanced with construction of its ground spare satellite, although we have heard rumors that work may have paused in anticipation of the ICO ruling, and (unless TerreStar gains a European license in the near future and decides to use the satellite there) we expect that construction will soon be formally suspended to save money, on the assumption that this will not impact TerreStar’s pending ATC application.
A future question for the FCC with regard to ATC “flexibility” may well relate to what level of satellite performance is needed to justify that an ATC service is truly “ancillary”. TerreStar and Skyterra have built very large and powerful (and expensive) satellites, in order to deliver voice and data connectivity to “standard” mobile devices without external antennas. ICO and Globalstar’s satellites are rather less capable (and cheaper), but if they are not required to deliver fully reliable handheld satellite voice services, then these satellites could also operate with “standard” mobile devices. For example, if the only test required is to complete a call from a handset on a tripod in an open field, with no head blockage or obstruction from trees etc, then a far less expensive satellite is required than if link margin is needed to overcome these obstructions and the call is made or received by a real person walking around with the phone.
Given that many hundreds of millions of dollars of satellite construction costs are at stake, it will be very interesting to see how the FCC ultimately decides this issue and therefore who has made the right call with their satellite design.
United has now joined the North American airlines signing up to fit the Aircell service for in-flight Internet connectivity. Similar to American, it is initially just installing the service on its P.S. business-oriented cross country flights between JFK and LAX/SFO (though American also includes some 767 flights to MIA). Undoubtedly this is a great boon for business travelers, and our experience of the service was excellent. However, to date it looks like overall usage levels are very low, since few leisure travelers are willing to pay $12.95 even for a five hour flight.
On a American SFO-JFK afternoon flight last October, we decided to walk the plane and count the number of users: the result was 8 out of 34 business and first passengers were using it, but only 2 out of about 110 economy passengers. I’m sure American is pleased with this – since the high revenue customers at the front of the plane are happy, but the amount of money flowing to Aircell is far from enough to pay for the network. We understand that to date Aircell has installed the equipment for free, so the only cost to the airline is the fuel to fly it around.
Based on the usage levels we saw, gross Aircell revenue is probably only ~$60K-$80K per plane per year, less even than the $100K seen by Connexion-by-Boeing back in 2006. Connexion had many of the same characteristics – giving away equipment, a high fixed cost network (in that case global satellite capacity leases rather than a national tower network), a large staff, and was also a great service for passengers and airlines. There are a few differences, most notably that the Connexion equipment was much heavier and more expensive than the Aircell terminals, but also that Aircell can supplement its passenger revenues with installations in the business jet market. However, Boeing ultimately decided it couldn’t afford to continue to run the service, as did Claircom, Airfone and others with their earlier voice services. In the current financial climate, we wonder if Aircell’s network will be able to avoid the same fate? Certainly they seem a long way from the prediction of 2000 equipped aircraft by the end of 2009 made by Aircell’s CEO last summer.